In the U.S. behavioral health insurance benefits have been characterized by fewer covered services and higher deductibles, copayments and coinsurance than medical care benefits. TRICARE’s new behavioral health parity rule implemented in October 2016 eliminated gaps between behavioral health and medical care benefits. This research examined a change in mental health among TRICARE beneficiaries attributable to the new parity rule. We analyzed observational data from the 2010-2020 Medical Expenditure Panel Survey, a well-established largest national survey that collects detailed information on verified health conditions and sociodemographic characteristics in a nationally representative sample of the non-institutionalized U.S. civilian population including TRICARE beneficiaries. The sample included non-elderly adults aged 18-64, including 1,166 TRICARE beneficiaries and 72,996 privately insured/uninsured individuals. Outcomes included perceived and validated mental health measures such as binary measure of self-assessed poor mental health and Mental Composite Summary (MCS) scores which are derived from the Short-Form 12 and range from 0 (worst) to 100 (best). We also constructed binary measures of serious mental illness (SMI) and moderate mental illness (MMI) using Kessler Psychological Distress Scale and its validated cut-points. We created an indicator for potential depression using the 2-item Patient Health Questionnaire (PHQ-2) and its validated cut-point. We employed a novel difference-in-difference-indifferences (DDD) model, a quasi-experimental statistical approach that can uncover causal impacts of a policy change based on observational data. We first compared pre-post changes in the outcomes for the policy intervention group of TRICARE beneficiaries with counterfactual trends from the civilian counterparts and when further differenced the pre-post/policy-control changes for individuals with serious behavioral health conditions and other individuals without behavioral health conditions. All estimates were adjusted for comprehensive demographic and socio-economic characteristics as well as various time trends. We confirmed that our sample satisfies the parallel trend assumption that underlies the validity of our analytic design. Estimates were survey-weighted and adjusted for a multistage complex survey design. Findings show that the implementation of new behavioral health parity rules implemented in TRICARE was associated with a 7.3-point increase in MCS scores (p=.003), a 25.5 percentage-point decrease in the probability of having SMI (p<.001), a 20.5 percentage-point decrease in the probability of having potential depression (p=.013), and a 9.7 percentage-point decrease in the probability of reporting poor mental health (p<.001) among TRICARE beneficiaries. The beneficial impacts of behavioral health parity grew in magnitude as the family income level decreases. Behavioral health parity was associated with 55.1, 35.1, 28.7, and 10.4 percentage-point decreases in the probability of having SMI among individuals living under poverty (p=.001), individuals with low family income (p<.001), individuals with moderate family income (p<.001), and individuals with high family income (p<.001), respectively. Similar patterns were discovered for MCS scores and the probabilities of having potential depression and self-assessed poor mental health. We conclude that the implementation of behavioral health parity overall led to significant improvement in mental health among TRICARE beneficiaries, especially for those in financial need.